Assertation error running example pytorch notebooks

[al3xsh]

Hi,

I get the following error when I try and run your example pytorch notebooks:

~/code/development_tools/anaconda3/envs/fastai/lib/python3.7/site-packages/torch/onnx/symbolic.py in adaptive_avg_pool2d(g, input, output_size)
    598 @parse_args('v', 'is')
    599 def adaptive_avg_pool2d(g, input, output_size):
--> 600     assert output_size == [1, 1], "Only output_size=[1, 1] is supported"
    601     return g.op("GlobalAveragePool", input)
    602 

AssertionError: Only output_size=[1, 1] is supported

I haven't modified anything, and I wonder if you have any idea what the source of this error is?

Regards,

Alex

[v1093]

Hi,

I get the following error when I try and run your example pytorch notebooks:

~/code/development_tools/anaconda3/envs/fastai/lib/python3.7/site-packages/torch/onnx/symbolic.py in adaptive_avg_pool2d(g, input, output_size)
    598 @parse_args('v', 'is')
    599 def adaptive_avg_pool2d(g, input, output_size):
--> 600     assert output_size == [1, 1], "Only output_size=[1, 1] is supported"
    601     return g.op("GlobalAveragePool", input)
    602 

AssertionError: Only output_size=[1, 1] is supported

I haven't modified anything, and I wonder if you have any idea what the source of this error is?

Regards,

Alex

For me, downgrading torchvision from 0.2.2 to 0.2.1 worked.
So basically, torchvision 0.2.2 has an AdaptiveAvgPool2d instead of AvgPool2d in AlexNet.
It was mentioned in PyTorch discussions page: https://discuss.pytorch.org/t/when-using-torch-onnx-export-errors-occurs-assertionerror-only-output-size-1-1-is-supported/38528/3